Representations of Imaginary, Nonexistent, or
Nonfigurative Objects
Representações de Objetos Imaginários, Não-existentes ou
Não-figurativos
Winfried Nöth
Interdisciplinary Center for Cultural Studies
University of Kassel – Germany
[email protected]
Abstract: According to the logical positivists, signs (words and pictures) of
imaginary beings have no referent (Goodman). The semiotic theory behind
this assumption is dualistic and Cartesian: signs vs. nonsigns as well as the
mental vs. the material world are in fundamental opposition. Peirce’s
semiotics is based on the premise of the sign as a mediator between such
opposites: signs do not refer to referents, they represent objects to a mind,
but the object of a sign can be existent or nonexistent, a feeling or an idea,
something merely imaginary or even something false. The paper examines
the nature of imaginary, nonexistent, and nonfigurative objects, such as
unicorns, phoenixes, or nonrepresentative paintings, and shows why verbal or visual representations of imaginary beings and even nonfigurative
paintings, which seem to represent nothing, are fully developed signs with
a variety of objects determining their representation.
Keywords: Representation. Object of the sign. Referent. Pictures. Charles S.
Peirce. Nelson Goodman.
Resumo: De acordo com os positivistas lógicos, signos (palavras e figuras) de
seres imaginários não têm referentes (Goodman). A teoria semiótica por trás
dessa assunção é dualista e cartesiana: signos e não-signos, assim como
também mundo mental e mundo material estão em franca oposição. A
semiótica de Peirce é baseada na premissa do signo como mediador entre
esses opostos: signos não se referem a referentes, eles representam objetos a uma
mente, mas o objeto de um signo pode ser existente ou não-existente, um sentimento ou uma idéia, algo meramente imaginário ou até mesmo falso. O
artigo examina a natureza de objetos imaginários, não-existentes e não figurativos, tais como unicórnios, fênixes ou pinturas não-representativas, mostrando por que representações verbais ou visuais de seres imaginários e até
mesmo pinturas não-representativas, que parecem não representar nada, são
signos completamente desenvolvidos com uma variedade de objetos que determina sua representação.
Palavras-chave: Representação. Objeto do signo. Referente. Figuras. Charles S.
Peirce. Nelson Goodman.
Cognitio, São Paulo, v. 7, n. 2, p. 277-291, jul./dez. 2006
277
Cognitio – Revista de Filosofia
1. Objects, the Object of the Sign, and the Referent
In everyday usage, the word object denotes a solid, visible, tangible, and inanimate
thing; the notion of a nonexistent or merely imaginary object must appear as a
contradiction in terms. In its philosophical origin, by contrast, the concept of object did
not really mean an object of material existence but rather a mental object. The scholastics,
who introduced the term to the vocabulary of Western philosophy, did not conceive of
an object as something that had to belong to physical reality; instead, an object was an
object of thought, knowledge, or cognition, and “to be objectively” meant “to be in the
mind” (cf. FERRATER MORA 1975: 311).
When Charles S. Peirce uses the concept of the object of a sign in the sense of
something “perceptible, or only imaginable, or even unimaginable in one sense” (CP
2.230, 1910), his concept of object is quite in accordance with the medieval usage of
the concept. In fact, Peirce’s usage of the term object was directly influenced by 13th
century scholastic terminology, where, as he reminds us, objectum meant “a creation of
the mind in its reaction with a more or less real something […] upon which cognition is
directed” (MS 693A: 33; PAPE 1996: 115).
Peirce has a very broad concept of the object of the sign (cf. SANTAELLA 1988,
1990, 1994; JOSWICK 1996; PAPE 1996). The object of the sign is certainly not a “thing,”
and it need not be an existent at all. The sign does not refer to, it represents its object.
Only genuinely indexical signs refer in the sense that they indicate their object, since
only their objects are “existent individuals (whether things or facts)” (CP 2.283, 1902);
other indices, symbols, and icons do not refer, they represent, for example “something
of a general nature” or even something “believed formerly to have existed or expected
to exist” (CP 2.232, 1910). Above all, the object of the sign is that information, knowledge,
or experience which an observer of the sign must have in order to be able to interpret
the sign. The study of signs (words or pictures) referring to “imaginary objects” is hence
a natural branch of the tree of semiotic inquiry. In Peirce’s semiotics, all signs have
objects, even the word and. The object of this symmetrical conjunction, for example, is
the idea as well as the practical experience of how entities (terms or things) are combined
to form a whole (cf. SAVAN 1994: 189).
In the reductionistic framework of 20th century positivist semantics of reference,
the concept of object (e.g., QUINE 1960), sometimes called the referent (Ogden and
Richards 1923), or denotatum (CARNAP 1942; GOODMAN 1978), became largely
restricted to real and material objects with the result that the study of words or pictures
referring to imaginary or otherwise nonexistent objects either had to be excluded from
analysis or resulted in analytical aporias left to be resolved in a dualistic semantics
distinguishing between reference to concrete objects (denotation) and sense as the
ideas associated with the sign (signification). The present paper intends to discuss some
of these aporias and to contrast the views of the logical positivists on the semiotic
dualism between existent vs. nonexistent objects to Peirce’s holistic theory of the object
of verbal and pictorial representation.
278
Cognitio, São Paulo, v. 7, n. 2, p. 277-291, jul./dez. 2006
Representations of Imaginary, Nonexistent, or Nonfigurative Objects
2. Goodman’s Cartesian and Positivist Views of Pictorial Representation
The distinction between sense and reference (or signification and denotation) drawn by
20th century logical semanticians, following FREGE (1892), has been used to account for
the difference between expressions such as horse or dog on the one hand and unicorn,
dog with three heads, or the first woman who landed on the moon on the other (cf.
KEMPSON 1977: 13-14). The former expressions have both sense, which can be described
by means of paraphrases or definitions, and reference or referents in the form of objects
which can be pointed out in the world of existing things. The latter expressions, by
contrast, which do not refer to anything that really exists, are without reference although
they have sense, since their meaning, despite their semantic anomalies, can be imagined,
paraphrased, or translated into another language.
This Fregean model of analysis implies a split theory of semantics according to
which there are expressions with, and expressions without, a fully developed semantics,
that is, expressions with both sense and reference, e.g., horse or dog, expressions with
sense but without reference, e.g., winged horse or dog with three heads, and finally
expressions with reference but without sense, such as proper names designating
individuals, e.g., George Washington. Ultimately, this semiotic split has a Cartesian
foundation, since it takes for granted that the cognitive world is divided into two, on the
one hand, a material world of perceptual objects which serve as referents of signs and
on the other hand a mental world of ideas which constitute the realm of sense or
signification. Below, we will see how Peirce’s anti-Cartesian semiotics is free of this
dualism, but first some of the aporias of the still prevailing Cartesian model will be
expounded with special reference to Nelson Goodman’s views of pictorial representation.
Goodman’s pictorial semiotics extends the positivist semanticians’ approach to
sense and reference in language to pictures, especially to paintings. In his paper “Routes
of reference,” Goodman draws the following parallels between verbal and pictorial signs
(in his terminology: “symbols”) to conclude that both verbal and pictorial signs of fictional
and imaginary objects are empty signs without referents:
Some names and descriptions and pictures – such as “Robinson Crusoe” or
“winged horse” or a unicorn-picture – denote nothing although each belongs to
a system along with other symbols that denote one or many things. To hypostatize
a realm of nonactual entities for these empty symbols to denote seems to me
pointless and confusing. When we speak of a picture as depicting a unicorn,
even though there are no unicorns to depict, what we are saying in effect is
rather that the picture is a unicorn-picture; we are saying not that the picture
denotes anything but rather that it is denoted by the term “unicorn-picture.”
And we can distinguish unicorn-pictures from centaur-pictures, as we distinguish
desks from tables, with no regards to their denoting anything. (GOODMAN
1981: 125)
Goodman’s argument is not only that words and pictures depicting imaginary objects
are empty signs (“symbols”) without denotation; with his parallelism between chairs,
desks and unicorn-pictures, he moreover suggests that the real semiotic function of the
empty sign of a unicorn-picture is to be the referent (denotatum) of another sign, the
verbal expression “unicorn-picture.” The logocentrism of this argument is blatant, and
the chain of referents thus established testifies to the priority given to verbal logic over
Cognitio, São Paulo, v. 7, n. 2, p. 277-291, jul./dez. 2006
279
Cognitio – Revista de Filosofia
of visual semiosis: the metasemiotic verbal expression unicorn-picture is recognized as
a fully developed symbol, while the object-semiotic picture of the unicorn to which it
refers and which gives rise to the idea metasemiotically expressed is reduced to the
function of an “empty” symbol.
Symbol, representation, denotation, and reference are the key terms in Goodman’s
semiotic vocabulary. A symbol, in Goodman’s philosophy of art and literature, can be
verbal or nonverbal, a word or a picture. In the terminology of Peirce’s general semiotics,
what Goodman calls “symbol” is a sign, a term which Goodman does not use once. A
representation, according to GOODMAN (1978: 20) is something that is a sign of, refers
to, stands for, or symbolizes, a denotatum. A painting which depicts nothing but a
composition of colors and forms, such as a Mondrian, is hence a “nonrepresentational
picture which says nothing, denotes nothing, pictures nothing.” Despite this claim that
such pictures neither denote, represent, nor “picture,” Goodman goes on to conclude
that nonrepresentational paintings are symbols and have a “referential function.” This
conclusion sounds enigmatic, for, how can something be a symbol, but “say nothing,”
and what can a picture refer to if it does not represent anything?
Goodman’s terms representation and denotation designate the relation of signs
or symbols to their object (denotatum), but Goodman’s positivist concept of the
denotatum is very narrow: a symbol can only denote or represent something that exists
in a world that is “external or extraneous” (ibid.: 60) to this symbol. The semiotics
behind this argument is based on a sharp dualistic distinction between semiotic
phenomena intrinsic to symbols and nonsemiotic phenomena extrinsic to them, i.e.,
between a semiotic and a nonsemiotic world. A symbol denotes or represents either
uniquely or generally. For example, “names and such pictures as individual and group
portraits denote uniquely, while predicates and such pictures as those in an ornithologist’s
guide denote generally” (ibid.: 103), but symbols of fictional or imaginary objects do not
denote at all. Hence, Goodman concludes that “painted or written portrayals of Don
Quixote, for example, do not denote Don Quixote – who is simply not there to be
denoted” (ibid.), or that “Bosch’s paintings of weird monsters, or the tapestry of a unicorn,
represent nothing; for there are no such monsters or demons or unicorns anywhere but
in such pictures or in verbal descriptions” (ibid.: 60). The “axe of dualism” (CP 3.570)
which Goodman swings with such reductionistic distinctions between the real and the
imaginary blocks his view to more subtle inquiries into the roots of imagination not only
in cultural traditions, but also in real life experience. For example, representations of
weird monsters are not only representations of nonexistent beings, they are also
transformed representations of the real fear which humans have experienced in the
face of really existing dangerous animals in the wild.
If imaginary beings can neither be represented nor denoted by a picture and
unicorn- or sphinx-pictures are hence nonrepresentational, nonfigurative paintings of
pure forms and colors are even clearer examples of nonrepresentational pictures,
according to Goodman. However, under these premises, the important distinction drawn
in the traditional semiotics of painting between figurative vs. nonfigurative pictures
(unicorn-pictures vs. monochrome pictures, e.g.) gets lost, and this seems to be
unsatisfactory even to Goodman himself, since he admits the necessity of introducing
an additional distinction between these two kinds of pictures: “Let’s agree,” he argues in
his comment on unicorn pictures (ibid.), “that such pictures, though they represent
nothing, are representational in character, hence symbolic […]. All the same, we must
280
Cognitio, São Paulo, v. 7, n. 2, p. 277-291, jul./dez. 2006
Representations of Imaginary, Nonexistent, or Nonfigurative Objects
note in passing that their being representational involves no representation of anything
outside them.” Goodman’s extended argument is that pictures of imaginary creatures, in
their figurative form, are indeed similar to pictures which represent real objects against
a ground, especially when the imaginary creatures are inserted in a figurative scene
such as a landscape. The words designating these creatures are like other words (and
therefore “representational in character”) since they are phonetically and morphologically
constructed like other verbal symbols, with the exception that they are devoid of
reference. To call unicorn-pictures “representational in character” is thus a concession to
their undeniable figurativity, but the argument that the picture of a unicorn represents
nothing “outside the representation” suffers from a blatant reductionism. Not only does
the picture of a unicorn represent iconically a white horse with a long straight horn
growing on its head but also does it represent iconically innumerable other unicorns well
known from the history of Western painting and mythology. All of these elements
represented in the unicorn picture represent certainly something from “outside” the
picture.
In sum, Goodman distinguishes three main modes of pictorial representation. The
first can be exemplified by a portrait of myself; it is a symbol which represents, denotes,
and refers to me. Let us call it a picture of an existent object. The second example is a
picture of Don Quixote or of a unicorn; both are symbols, have reference, represent
nothing, denote nothing, but are “representational in character.” Let us call such pictorial
representations pictures of imaginary (or fictional) objects. The third kind of picture
may be exemplified with a Mondrian or a monochrome painting; it is a symbol which
has reference but is not representational in character, does not represent, and does not
denote. Pictures of this kind are often called abstract or nonrepresentational, but the
former term is inappropriate because it still suggests figurativity (one which merely
“abstracts” from details), and the latter is inappropriate in this context, since Goodman
describes pictures of imaginary objects as equally nonrepresentational (whereas according
to Peirce, as we shall see below, every picture “represents”). Table 1 gives a summary
of the criteria and examples discussed so far.
Goodman’s criteria
is symbol and refers;
has referential function
is representational
in character
represents and denotes
(in “external” existence)
Picture of existent objects;
photographic portrait
+
+
+
Picture of imaginary (or fictional)
objects; unicorn, Don Quixote
+
+
–
Nonfigurative picture;
monochrome painting
+
–
–
Type of picture; example
Table 1. Three kinds of pictures according to Goodman
Let us now try to solve the enigma why a picture which does not represent can
nevertheless be a symbol. Goodman develops his argument with examples from pictures
of the third kind, nonfigurative pictures which “represent nothing” and are not
“representative in character” either. A Mondrian, he says, is a symbol because it
“exemplifies,” “stands as a sample of,” “makes manifest, selects, focuses upon, exhibits,
[or] heightens in our consciousness” properties which it possesses itself (ibid.: 65).
Cognitio, São Paulo, v. 7, n. 2, p. 277-291, jul./dez. 2006
281
Cognitio – Revista de Filosofia
Describing nonfigurative paintings in these terms, Goodman evidently means pictures
that are signs insofar as they refer (and therefore “have reference”) to themselves. In
other words, his topic is self-referential pictures, a term which Goodman does not use,
but with which he could not disagree, since, in his words, “exemplification is surely to
symbolize […, and it] is a form of reference. A work of art, however free of reference
and expression, is still a symbol even though what it symbolizes be not things or people
or feelings but certain patterns of shape, color, texture that it shows forth” (ibid.).
Reference is hence a category in Goodman’s semiotics besides representation
and denotation; it is that criterion which makes a visual phenomenon a symbol. Whether
they represent or not, my portrait, the unicorn, and the Mondrian pictures all evince
reference and are symbols for this reason. The difference between my portrait and the
Mondrian picture is that the former refers to a person who is “extrinsic” to the picture,
living in a world other than the world of pictures, whereas the unicorn and the Mondrian
pictures refer to something which is “intrinsic” to the picture.
3. Peirce on Signs of Imaginary Objects
Let us now turn to Peirce’s semiotic answers to the above discussed enigmas of pictorial
semiosis. For the sake of brevity we must omit Peirce’s analyses of pictures of really
existent objects, which have been much discussed in the context of the theory of
pictorial iconicity and indexicality (cf. NÖTH 2000, 2003; SANTAELLA; NÖTH 2003).
Instead, we will restrict the discussion to pictures of imaginary and nonfigurative objects.
According to Peirce, every picture is a sign. In this respect, Peirce and Goodman
agree, except for their terminology (Goodman’s term is “symbol”). To define pictures in
general as signs (or symbols) is by no means a generally accepted premise; aestheticians
based in phenomenology have often claimed that pictures are not signs but phenomena
sui generis (cf. NÖTH 2005).
Peirce did not elaborate a semiotics of painting let alone of nonfigurative art, of
which he could hardly have any knowledge until his death in 1914. However, his general theory of signs and his many comments on pictures, photos, images, and icons
provide a unified framework for pictorial semiotics applicable to figurative and
nonfigurative pictures as well as to pictures of imaginary beings. His theory of the object
of the pictorial sign dispenses with many of the aporias of the positivists whose dualist
theory of sense and reference is only fully applicable to pictures representing “really”
existing objects.
The most important difference in comparison with the positivists is that Peirce
rejects the Cartesian distinctions between sense and reference on the one hand and
between the sign and the nonsemiotic world “extraneous” to the sign on the other. The
object of a sign can be a “real” thing of practical experience, a mere sign, a mental
representation, an abstract notion, or an idea of something purely imaginary. Peirce says
nothing about the “reality” of the object at all and describes it as something “perceptible,
or only imaginable or even unimaginable in one sense” (CP 2.230, 1910) and even goes
so far as to speculate that “perhaps the Object is altogether fictive” (CP 8.314, 1909).
Hence, imaginary or fictional beings, such as a unicorn or Don Quixote are naturally
among the objects of verbal or pictorial signs. By conceiving of the object of the sign as
282
Cognitio, São Paulo, v. 7, n. 2, p. 277-291, jul./dez. 2006
Representations of Imaginary, Nonexistent, or Nonfigurative Objects
something that may be a sign itself, but a sign which precedes the actual sign, Peirce
manages to avoid many of the pitfalls and semiotic aporias of the positivists.
The distinction between representational and nonrepresentational pictures would
have been a contradiction in terms for Peirce, since in his semiotics, sign and representation
are quasi-synonyms (cf. NÖTH 1997). Every sign “represents” (instead of “refers to”) an
object. In contrast to the positivist position, according to which only “real” in the sense
of existent objects can be represented, the object of a pictorial sign, according to Peirce,
can be another sign, another picture, the mere memory of the experience of an object,
a mental image, the sensation of a color, or, in the case of synaesthesia, even the sensation
of a nonvisual sense impression, a smell, a sound, or a taste.
This is why signs can represent and even denote merely fictional beings, for
example, a phoenix or a centaur, two examples discussed by Peirce. A picture of a
phoenix has an object, says Peirce in 1910, “for although no phoenix really exists, real
descriptions of the phoenix are well known” (CP 2.261), and since the object of the sign
is that which is known about it, the sign of a phoenix has evidently an object. But what
is the nature of such an object that does not “really exist”?
Peirce distinguishes two kinds of object, the immediate and the dynamical object.
The dynamical object is the reality behind the sign, “the Object outside of the Sign” (EP
480, 1908). It belongs to a reality which is independent of its representation in the sign,
a reality to which the interpreter of the sign has no full access, since only “unlimited and
final study could reveal” its nature (CP 8.181, 1903) at the end of an endless series of its
representations (CP 1.339, 1893). Of this object we have to assume that it “by some
means contrives to determine the Sign to its Representation” (CP 4.536, 1906), which
means that the dynamical object must precede the sign which it determines to its
representation (see below).
The immediate object, by contrast, is “the object as the sign represents it” (CP
8.343, 1910) and “whose Being is thus dependent upon the Representation of it in the
Sign” (CP 4.536, 1906), which means that it is an object only incompletely presented by
the sign (CP 4.536, 1906; CP 8.183, 1903). This description of an incomplete presence
of the dynamical object in the immediate object fits well with another requirement,
familiarity or presupposed knowledge. The immediate object is that which we know
about the object, and familiarity and previous knowledge can never be complete
knowledge of the object. The argument that the object “is itself of the nature of a sign
or thought” (CP 3.538, 1903) is equally compatible with the notion of the “object of the
sign as the sign represents it.” As an incompletely represented object, the immediate
object may be vague, opaque, false, erroneous, or otherwise affected by prejudice, bias,
or cultural ways of seeing the things. The immediate object contains an index, a “hint”
indicating the dynamical object to which it corresponds (EP 480, 1908).
The difference between the immediate and the dynamical object seems relatively
clear in signs referring to physical facts and our knowledge of them. The word “electricity,”
for example, had a different and less evolved immediate object present to Benjamin
Franklin’s mind. The knowledge which physicists of our times have about the same
phenomenon has grown, and the immediate object of the sign has changed over the
centuries. The dynamical object, by contrast, has remained unchanged in the course of
time, since this object is the sum total of all physical facts which cause experts and lay
persons likewise to denote electricity by its verbal sign (cf. SANTAELLA 1995: 53-66),
and it will remain the same if the future should reveal more or other facts about the
Cognitio, São Paulo, v. 7, n. 2, p. 277-291, jul./dez. 2006
283
Cognitio – Revista de Filosofia
physical processes involved with the object of this sign.
Another term by which Peirce used to designate the dynamical object was real
object, and if the dynamical object of the sign is the reality behind the sign, the example
of the word “electricity” exemplifies well the idea that the real object is something
inaccessible but unchangeable, determining each individual sign of it while more and
more knowledge is being gathered about its reality. The concept of reality which Peirce
proposes reveals the nature of the dynamical or real object, too. “I define real,” explains
Peirce in 1906 (CP 6.495) “as that which holds its characters on such a tenure that it
makes not the slightest difference what any man or men may have thought them to be,
or ever will have thought them to be, here using thought to include, imagining, opining,
and willing […]; but the real thing’s characters will remain absolutely untouched.”
But what about the dynamical object of thought-signs, fictions, myths, and mere
imaginations? Can ideas as dynamical objects of signs “remain absolutely untouched” by
the thoughts people have had about them? What is the dynamical object of an imaginary
being in a picture which represents something that does not exist? Since the dynamical
object of a sign is that which determines the sign to its representation, the dynamical
objects of myths, fictions, ideologies, or imaginations are evidently to be found in the
system of ideas, myths and imaginations rooted in a culture insofar as it precedes a
particular representation in a specific sign, and the immediate object of such a cultural
representation is the imaginary or fictional object as it presents itself in an incomplete
way to the mind considering the specific sign.
Peirce gives the example of a picture of a phoenix (MS 318: 41, cf. PAPE 1996:
109). Although it represents a fictional being, he argues, it is certainly a sign, not only
with an immediate but also with a dynamical object. According to the above premises,
its immediate object is our familiarity with the image of a phoenix, which is not only
rooted in some usually incomplete knowledge of ancient mythology but also in our
knowledge of the real world of birds and fire, life and death. According to ancient Greek
mythology, a phoenix is a magic bird born from fire. The dynamical object of this sign is
hence the ancient mythological construct which has determined and will determine all
past and future representations of phoenixes. It is the sum total of what has been
known about these magical birds. However, the complex mental construct of this (mental) dynamical object does not only consist of magical and unreal archaic imaginations.
Insofar as ideas such as “fire,” “large bird,” or “resurrection (after a defeat)” are part of
the idea of a phoenix, fragments of present-day real world experience are part of the
cultural construct, too.
In 1908, Peirce calls the dynamical object the “really efficient but not immediately
present Object” (CP 8.342). The archaic (and hence not “immediately present”) complex
of cultural and natural ideas, images, and anatomical details associated with the figure of
a phoenix which have determined (have been “really efficient” with respect to)
innumerable of its narrative or pictorial representations since antiquity are their dynamical
objects in this sense; in their complexity and our historical distance from its origin and
evolution, these objects are never fully present to the mind of an observer of an individual picture of a phoenix. What is present and plays a role in the viewer’s understanding
of the picture of a phoenix is the immediate object, which consists of the observer’s
partial knowledge based on an incomplete familiarity with the dynamical or real object.
There are always further details from the broader horizon of culture and nature, the
dynamical object of the picture, of which the viewer has insufficient knowledge or
284
Cognitio, São Paulo, v. 7, n. 2, p. 277-291, jul./dez. 2006
Representations of Imaginary, Nonexistent, or Nonfigurative Objects
awareness but which are likely to determine and at the same time limit the possibilities
of future interpretations.
Peirce’s typology of possible dynamical objects is triadic, as all of his categories
are. In 1908, he writes that the sign may represent its dynamical object as a mere
possibility, an existent, or a necessity. Only the second of these categories is applicable
to representations of “real” existents and facts. His examples of signs determined in this
way by their dynamical object are a barometer and “a written narrative of any series of
events.” Fictional narratives, myths, or imaginary creations, by contrast, due to their lack
of realism, are determined by mere possibilities, since “a Possible can determine nothing
but a possible” (EP 480-81). However, insofar as they also represent historical contexts,
as Shakespeare’s Hamlet, they may also be determined by the universes of existents
and necessities, such as historical facts which cannot be denied.
Tellingly, in 1903, Peirce exemplifies his semiotic category of the iconic sign not
with the example of a picture of some existent object, but with the example of a
representation of the statue of a centaur. Peirce’s argument why such a statue is an icon
although it represents a merely imaginary being is the following: “The statue of a centaur
is not, it is true, a representamen if there be no such thing as a centaur. Still, if it represents
a centaur, it is by virtue of its shape; and this shape it will have, just as much, whether
there be a centaur or not” (CP 5.73, 1903). Hence, the statue of the centaur is an iconic
sign because its features correspond to those of other representations of a centaur
which are equally determined by the cultural prototype of this mythological creature
with its mixture of features of a human and of an animal being.
One of the aspects of the dynamical object of a mythical creature is its fictional
anatomy. A centaur, for example, is a fictional creature from Greek mythology with the
head, chest, and arms of a man and the body and legs of a horse. These anatomical
characteristics are features of the dynamical object which determine any pictorial
representation of a centaur. For example, with reference to these features we can decide that an appropriate picture of a centaur cannot have a bird’s head. A representation
which has the typical features of a centaur is therefore an icon of the dynamical object
of this ancient mythological creature, although no real creature of this kind exists in
nature. Iconicity, too, does not necessarily require correspondence with something really
existing; it is rather based on the correspondence of the features of one mental image
with those of another.
4. Thought-signs, the Object in the Mind, True and False Representations
The object of a sign need not exist in reality, but in order to understand a sign, its
observer must be familiar with this object, which is not contradictory, since familiarity
does not presuppose existence; we are not only familiar with facts, but also with ideas,
fictional beings, errors, or even lies. Familiarity presupposed, the object of the signs is a
feeling, experience, or cognition which is antecedent to the sign. Nothing can be a sign
if the object is completely unknown, although the object may be nonexistent, imaginary,
or even false (e.g., CP 2.230, 1910).
Since 1868, Peirce describes thoughts as examples of signs, which he calls thoughtsigns (CP 5.283, W 223; cf. SHORT 2004: 215). In 1896, he argues that the first of a
Cognitio, São Paulo, v. 7, n. 2, p. 277-291, jul./dez. 2006
285
Cognitio – Revista de Filosofia
representation can be exterior or interior to the interpreting mind and that there can be
an “outward or inward” mediation “between an object and an interpreting thought” (CP
1.480; 1896). Signs can hence be mental representations, and mental representations
are signs, according to Peirce. As new as this projection of the mental into the chain of
signifiers may seem, its roots are in medieval semiotics (cf. TABARRONI 1989).
The thesis that silent and invisible mental representations are signs seems
counterintuitive from the commonsensical point of view according to which signs serve
as instruments in human communication and must therefore be perceptible to an
addressee. However, the purpose of a sign, according to Peirce is not necessarily to be
an instrument for the communication of ideas to others. The purpose of a sign is rather
to “be interpreted in another sign” (MS 1476, 1904 [in: PEIRCE 1991]), and such
interpretations of the sign in a subsequent sign begin in the flow of ideas in a mind
addressing itself only to itself (CP 5.284, 1868).
The flow of ideas is a flow of signs in which “every thought-sign is translated or
interpreted in a subsequent one” (ibid.) so that each actual thought has its object in an
antecedent thought-sign and its interpretant is the subsequent thought in the flow of
thought-sign. With this account of semiosis in the flow of ideas, Peirce gives a radical
solution to the structuralists’ alleged dilemma of the semiotic abyss between the sign
and its object (the signifier and the signified). By projecting the object of the sign into
the chain of signifiers (alias representamens), Peirce semioticizes the object of a sign in
the most radical way. What is a thought-sign in the present moment will be one of the
objects of the subsequent thought-sign which it triggers. However, the difference between
the sign and its object in this flow of thought-signs is not merely one of the sequence in
time, but also one of determination of the subsequent sign by the antecedent object and
in the reverse perspective one of representation of the antecedent object in the
subsequent sign.
In 1903, Peirce sets up the requirement that any object of a sign be a thought:
“Every sign stands for an object independent of itself; but it can only be a sign of that
object in so far as that object is itself of the nature of a sign or thought” (CP 3.538, 1903).
The argument does not mean that Peirce turned mentalist, but it has to do with the
requirement of familiarity with the object, which can only be a familiarity in a mind. In
1871, Peirce specifies: “To say that an object is in the mind is only a metaphorical way
of saying that it stands to the intellect in the relation of known to knower” (CP 8.18).
The sign “can only represent the Object and tell about it,” but it “cannot furnish
acquaintance with or recognition of that Object” (CP 2.231, 1910), since it would be a
contradiction in terms to furnish acquaintance with something whose acquaintance is
presupposed in the first place. On the other hand, the sign does not only represent
something with which we are already acquainted; at the same time, its function is to
“convey some further information concerning” the object (ibid.).
There are two opposite directions of relation between the sign and its object; the
sign represents its object, and it is at the same time logically determined by it
(PARMENTIER 1985; PAPE 1996). Determination means that the object as the antecedent
information presupposed by the sign, the necessary familiarity with it, the so-called
collateral knowledge of what is represented in the sign, exerts a semiotic influence in
the way the sign is interpreted.
How can a sign be determined by something that does not exist, is merely
imaginary, or even erroneous? Peirce offers an example from literary fiction, the statement,
286
Cognitio, São Paulo, v. 7, n. 2, p. 277-291, jul./dez. 2006
Representations of Imaginary, Nonexistent, or Nonfigurative Objects
“Hamlet was insane.” Even though this dicent sign is about a merely fictional man, it has
an object, namely that which we know about Hamlet as the protagonist of Shakespeare’s
drama before we hear the sign. The object is the information necessary to understand
this sign; Peirce paraphrases it as the “Universe of Shakespeare’s Creation so far as it is
determined by Hamlet being a part of it” (ibid.). In fact, the object of a fictional sign is
a special kind of object, one about which Peirce says that it is “created by the sign” itself
(CP 8.178, 1903), but at the same time, it is not completely created by the statement
about Hamlet, since to understand the sentence about Hamlet’s insanity, we must have
some collateral knowledge of what insanity is, what a young man is like etc.
How even false or erroneous signs can represent an object is the topic of an
illustration given by Peirce in 1903 concerning the object of the historically false proposition
“Napoleon was a lethargic man” (CP 8.178). According to Peirce’s analysis, this sentence
has several objects determining the mind of the one who utters it. The first object is
Napoleon, the historical figure, of which we must have heard if we want to understand
the message. The second partial object is lethargy, since the sentence cannot “convey
its meaning unless collateral experience has taught its Interpreter what Lethargy is”
(ibid.). The truth missing in this dicisign, namely the fact that Napoleon was by no
means a lethargic man, is equally one of the objects of this sign, since truth, as Peirce
defines it in 1904 and 1906, is “the conformity of a sign to its object” or simply “the
object of the sign” (EP 380, 304) and determines its interpretation in its resistance to its
falseness in the long run.
With Peirce, we must conclude that any false sign has at least two dynamical
objects, which determine it as if in a dispute, as he writes in 1906: “So, then, a sign, in
order to fulfill its office, to actualize its potency, must be compelled by its object. This is
evidently the reason of the dichotomy of the true and the false. For it takes two to make
a quarrel, and a compulsion involves as large a dose of quarrel as is requisite to make it
quite impossible that there should be compulsion without resistance” (CP 5.554).
5. Peircean Foundations of a Semiotics of Nonfigurative Pictures
If pictures are signs, and all signs represent an object, what does a nonfigurative picture
represent, and what is the object of such a picture? The answers to these questions that
can be derived from Peirce’s semiotics are several, and each will focus on a different
aspect of pictorial representation.
The first can be found in Peirce’s theory of iconicity, more specifically Peirce’s
theory of the pure icon. This is Peirce’s semiotic concept that comes closest to the selfreferential sign postulated in semiotic aesthetics and also in Goodman’s theory of
nonrepresentational art. In contrast to the hypoicon, which is a sign by similarity, the
pure sign is a sign which is so much like its object that it seems indistinguishable from it.
According to a manuscript of 1903, it is a sign that “does not draw any distinction
between itself and its object […,] an affair of suchness only” (CP 5.74, 1903), or “a
representamen of what it represents […] by virtue of […] characters which belong to it
in itself as a sensible object, and which it would possess just the same were there no
object in nature that it resembled, and though it never were interpreted as a sign” (CP
4.447, 1903).
Cognitio, São Paulo, v. 7, n. 2, p. 277-291, jul./dez. 2006
287
Cognitio – Revista de Filosofia
Peirce illustrates the notion of a pure icon with the example of a painting into
which the viewer becomes so immersed that he or she begins to see the representation
becoming indistinguishable from what it represents: “In contemplating a painting, there
is a moment when we lose the consciousness that it is not the thing, the distinction of
the real and the copy disappears, and it is for the moment a pure dream […]. At that
moment we are contemplating a [pure] icon” (CP 3.362, 1885). In the moment of
contemplative immersion described here, the picture no longer fulfills its usual sign
function of representing something else. Instead, it represents only itself and has thus
become a self-representing or self-referential picture
It is true that Peirce could not have meant a nonrepresentational picture in this
quote of 1885, but as far as the object relation is concerned, Peirce’s description of a
picture that becomes indistinguishable from its object to the contemplating mind is
rather congruent with Goodman’s interpretation of nonrepresentational pictures as symbols
referring to qualities “intrinsic” only to themselves. What is different in the two descriptions
of aesthetic perception apart from the terminology is that Goodman speaks of the
picture that symbolizes by showing forth its own “patterns of shape, color, texture,”
whereas Peirce describes a picture that becomes a matter of “mere suchness” while it
merges with its object in the process of contemplation. Goodman apparently conceives
of the self-referential nature of a nonfigurative painting that “shows forth” its own qualities
as a matter of indexical semiosis, whereas Peirce conceives pictorial self-reference as a
matter of pure iconicity. Below, we will see that, in fact, both aspects of pictorial semiosis
can coexist in one and the same painting.
Nonrepresentational paintings are complex signs, and their aesthetic features cannot
be reduced to their referring to their own qualities, which is a general description equally
valid for all nonfigurative works of art. In addition to self-reference, other sign relations
are relevant to the analysis of nonrepresentational art. Let us consider the example of
Kazimir Malevich’s painting Red Square and Black Square of 1914 or 1915 (see for
example http://www.artchive.com/artchive/M/malevich/blk_red.jpg.html). It shows, as
the title says, nothing but a red and a black square on a white ground, and in this radical
reduction of figures and forms, the painting seems to be a prototypical example of a
self-referential sign, a pure icon, that represents nothing but its own qualities. Being an
approximation to pure iconicity, the painting, like all icons, is a phenomenon of firstness,
the category of mere possibility and of suchness without any relation to anything else.
In their abstraction, the red and black squares offer indeed an unlimited potential of
representing real things in addition to the self-referential showing of its own qualities.
The pictorial sign is referentially open to the possibility that the two squares might not
only be red and black squares but also represent an indefinitely large number of red and
black square forms in the world of existing objects. For example, it could be the
representation of two square pieces of red and black paper on top of a sheet of white
paper used in a classroom to illustrate the geometry of squares. To the degree that the
possibilities of representation are unlimited and open, the pictorial sign is a pure icon, a
sign about which Peirce wrote that it “can convey no positive or factual information; for
it affords no assurance that there is any such thing in nature” (CP 4.447, 1903).
In addition to its self-referential character and its openness to many hypoiconic
interpretations, there are other respects, in which the semiotic potential of Malevich’s
painting is more determined in its message. On the one hand, it is not completely
without “positive or factual” information; on the other hand, it is not utterly devoid of
288
Cognitio, São Paulo, v. 7, n. 2, p. 277-291, jul./dez. 2006
Representations of Imaginary, Nonexistent, or Nonfigurative Objects
some general meaning. To the degree that the painting evinces factual information, it is
an index, to the degree that it conveys some general message, it is a symbol.
Malevich’s painting is not complete without its title Red Square and Black Square
which adds a verbal interpretation to the visual signs. In contrast to the equally possible
alternative title with articles, A Red Square and a Black Square, at least in the English
translation of the title, which would be indexical, pointing to the specific squares of this
particular picture, the title without articles suggests generality in its lack of any reference
to the specific pair of squares shown in this painting. Generality is a feature of symbolic
signs. Thus, insofar as the title suggests the more general idea of a red and a black
square, the two squares evince an element of symbolicity, too.
In other respects, the picture also evinces indexicality. For example, it conveys
positive information insofar as the two squares are precisely determined in their size,
shape and chromatic quality. Furthermore, the geometrical and chromatic forms constitute
oppositions, such as the ones between small and large, chromatic and nonchromatic
forms in horizontal or nonhorizontal, straight or oblique orientation. The resulting
composition constitutes a net of internal relations consisting of indexical references from
element to element: the chromatic red, a color that has a certain inherent indexicality
insofar as it is the natural color of alarm, points to its nonchromatic black counterpart; the
large square draws attention to the small square and its different size; the black square
arranged in parallel with the outlines of the rectangular frame draws attention to the
obliquely arranged red square, creating, by its formal opposition still another indexical
tension in the pictorial equilibrium.
The title Red Square and Black Square, in addition to its aspect of symbolicity
discussed above, suggests also an indexical reading of the painting. Apart from the
indexicality inherent in its being the name of an individual object, which makes any title
an index of the picture it refers to, it is indexical because it selects from the three colors
of the picture, red, white, and black, only two, thus focusing the viewers’ eyes selectively
to the red and the black squares in the picture. Furthermore, it also focuses selectively
insofar as it abstracts from the size and position of the square by not mentioning this
detail.
Malevich’s picture with the red and black squares has two different titles, the
second being Suprematist Composition. This title is indexical, too, in its selection of the
name of a style of painting and its identification of the painting as an example of this
school of art. Even without this second title the painting, like any painting characteristic
of a historical style, is an indexical sign of this style.
Let us conclude these considerations on the objects of nonfigurative paintings
with a second look at the topic of pictorial self-reference and pure iconicity. No real
painting can be absolutely self-referential and a pure icon in every respect. Only
approximations to pure iconicity are possible. To the degree that nonrepresentational
paintings are self-referential in their lack of figurative reference to the world of objects,
Malevich’s painting seemed to be an approximation to pure iconicity with its lack of
figurative referents. However, to restrict our attention to the aspect of the mere suchness
in the contemplation of the picture, as the contemplator described above by Peirce did,
who, as if in a pure dream, lost the consciousness of the fact that the painting was not
the object represented by it, would have obscured our view of many other semiotically
relevant features of this picture. As works of art, nonrepresentational paintings are complex
signs. The plurality of descriptions, interpretations, and categorizations to which they
Cognitio, São Paulo, v. 7, n. 2, p. 277-291, jul./dez. 2006
289
Cognitio – Revista de Filosofia
give rise testify to their semiotic complexity. After all, pictures which refer to nothing
but to themselves cannot become the object of critical and aesthetic discourse about
these pictures. Our analysis brought forth many indexical and symbolic features of
Malevich’s suprematist composition. Peirce has always emphasized that signs are rarely
of one kind only and has even claimed that “the most perfect of signs are those in which
the iconic, indicative, and symbolic characters are blended as equally as possible” (CP
4.448, 1903). The multiplicity of signs of all three kinds which we found in Malevich’s
suprematist composition may, in the end, testify to its aesthetic perfection.
References
CARNAP, Rudolf (1975 [1942/43]). Introduction to Semantics and Formalization of
Logic. Cambridge, MA: Harvard University Press.
FERRATER MORA, José (1975). Diccionario de filosofía, tomo 2. Buenos Aires:
Sudamericana.
FREGE, Gottlob (1975 [1892]). “Über Sinn und Bedeutung.” In: Funktion, Begriff und
Bedeutung. Göttingen: Vandenhoek. p. 40-65
GOODMAN, Nelson (1978). Ways of Worldmaking. Indianapolis: Hackett.
________ (1981). “Routes of reference.” Critical Inquiry, v. 8, n.1, p. 121-32.
JOSWICK, Hugh (1996). “The object of semeiotic.” In: Peirce’s Doctrine of Signs, V.
Eds.: M. Colapietro; T. M. Olshewsky. Berlin: Mouton de Gruyter. p. 93-102.
KEMPSON, Ruth M. (1977). Semantic Theory. Cambridge: University Press.
NÖTH, Winfried (1997). Representation in semiotics and in computer science.
Semiotica, v. 115, p. 203-15.
________ (2000). Handbuch der Semiotik. 2nd ed. Stuttgart: Metzler.
________ (2003). “Semiotic foundations of the study of pictures.” ShmeiwtikÞ: Sign
Systems Studies, Tartu, v. 31, n.2, p. 377-92.
________ (2005). “Warum Bilder Zeichen sind.” In: Bild-Zeichen: Perspektiven einer
Wissenschaft vom Bild. Ed.: S. Majetschak. München: Fink. p. 49-6.
OGDEN, Charles K.; RICHARDS, Ivor A. (1923). The Meaning of Meaning. New York:
Harcourt.
PAPE, Helmut (1996). “Object and final cause in Peirce’s semeiotic.” In: Peirce’s
Doctrine of Signs. Eds.: V. M. Colapietro; T. M. Olshewsky. Berlin: Mouton de Gruyter.
p. 103-117.
PARMENTIER, Richard J. (1985). “Signs’ place in medias res: Peirce’s concept of
290
Cognitio, São Paulo, v. 7, n. 2, p. 277-291, jul./dez. 2006
Representations of Imaginary, Nonexistent, or Nonfigurative Objects
semiotic mediation.” In: Semiotic Mediation. Eds.: E. Mertz; R. J. Parmentier. Orlando,
Fla.: Academic Press. p. 23-48.
PEIRCE, Charles Sanders (1931-58). Collected Papers. Eds.: C. Hartshorne; P. Weiss (v. IVI); A. W. Burks (v. VII-VIII). Cambridge, MA: Harvard University Press. (Quoted as CP)
________ (1963-66). The Charles S. Peirce Papers. Cambridge, MA: Houghton Library
of University Microproduction. (30 reels, microfilm edition.) (Quoted as MS)
________ (1991). Peirce on Signs. Ed.: J. Hoopes. Chapel Hill: University of North
Carolina Press. (Contains MS 1476)
________ (1998). The Essential Peirce, v. 2. Ed.: Peirce Edition Project. Bloomington:
Indiana University Press. (Quoted as EP)
QUINE, Willard v. O. (1976 [1960]). Word and Object. Cambridge, MA: MIT Press.
SANTAELLA, Lucia (1988). “Charles Sanders Peirce’s object (of the sign).” Versus v.
49, p. 53-8.
________ (1990). “An introduction to Peirce’s concept of object.” Cruzeiro
semiótico, Porto, v. 13, p. 75-82.
________ (1994). “Why things are objects of the sign.” Cruzeiro semiótico, Porto, v.
18, p. 37-46.
________ (1995). A teoria geral dos signos. São Paulo: Ática.
SANTAELLA, Lucia; NÖTH, Winfried (2003). Imagen: Comunicación, semiótica y
medios. Trad.: Roque Graciano. Kassel & Barcelona: Reichenberger.
SAVAN, David (1994). “C. S. Peirce and American semiotics.” In: The Peirce Seminar
Papers, 2. Eds.: M. Shapiro; M. Haley. Providence, RI: Berghahn. p. 179-208.
SHORT, Thomas L. (2004). “The development of Peirce’s theory of signs.” In: The
Cambridge Companion to Peirce. Ed.: Cheryl Mysak. Cambridge: University Press. p.
214-40.
TABARRONI, Andrea (1989). “Mental signs and the theory of representation in
Ockham.” In: On the Medieval Theory of Signs. Eds.: U. Eco; C. Marmo. Amsterdam:
Benjamins. p. 195-224.
Cognitio, São Paulo, v. 7, n. 2, p. 277-291, jul./dez. 2006
291